Fig 1 - available via license: Creative Commons Attribution 4.0 International
Content may be subject to copyright.
ROC-curves with the optimal cut-off value and associated specificity and sensitivity between brackets, for the first model a, containing the 6 MooMonitor+ behavioral metrics, the second model b, containing lactation number and DIM data, for the third model c, consisting of the 6 MooMonitor+ behavioral metrics and lactation and DIM and for the fourth model d, consisting of the 6 MooMonitor+ behavioral metrics, lactation, DIM, live weight and milk production data
Source publication
Background
Although visual locomotion scoring is inexpensive and simplistic, it is also time consuming and subjective. Automated lameness detection methods have been developed to replace the visual locomotion scoring and aid in early and accurate detection. Several types of sensors are measuring traits such as activity, lying behavior or temperatur...
Contexts in source publication
Context 1
... 1979 The first model consisted of MooMonitor + behavioral metrics. The second model contained data on lactation and DIM, the third model added the behavioral metrics to that, and the fourth model added live weight and production data on top of the data of the third model resulted to be 0.5, the associated specificity 0.52 and the sensitivity 0.66 (Fig. 1a). The second model achieved an optimal cut-off of 0.42, with a specificity of 0.79 and sensitivity of 0.73 (Fig. 1b). The optimal cut-off of the third model was 0.39, the specificity 0.83 and the associated sensitivity was 0.72 (Fig. 1c). For the fourth model the optimal cut-off was 0.54, the associated specificity 0.77 and sensitivity ...
Context 2
... DIM, the third model added the behavioral metrics to that, and the fourth model added live weight and production data on top of the data of the third model resulted to be 0.5, the associated specificity 0.52 and the sensitivity 0.66 (Fig. 1a). The second model achieved an optimal cut-off of 0.42, with a specificity of 0.79 and sensitivity of 0.73 (Fig. 1b). The optimal cut-off of the third model was 0.39, the specificity 0.83 and the associated sensitivity was 0.72 (Fig. 1c). For the fourth model the optimal cut-off was 0.54, the associated specificity 0.77 and sensitivity 0.80 (Fig. 1d). The accuracy of the four models was additionally checked using calibration plots (Fig. 2). The ...
Context 3
... top of the data of the third model resulted to be 0.5, the associated specificity 0.52 and the sensitivity 0.66 (Fig. 1a). The second model achieved an optimal cut-off of 0.42, with a specificity of 0.79 and sensitivity of 0.73 (Fig. 1b). The optimal cut-off of the third model was 0.39, the specificity 0.83 and the associated sensitivity was 0.72 (Fig. 1c). For the fourth model the optimal cut-off was 0.54, the associated specificity 0.77 and sensitivity 0.80 (Fig. 1d). The accuracy of the four models was additionally checked using calibration plots (Fig. 2). The calibration plots for the third and fourth model show good calibration, since most of the points were close to the straight ...
Context 4
... 1a). The second model achieved an optimal cut-off of 0.42, with a specificity of 0.79 and sensitivity of 0.73 (Fig. 1b). The optimal cut-off of the third model was 0.39, the specificity 0.83 and the associated sensitivity was 0.72 (Fig. 1c). For the fourth model the optimal cut-off was 0.54, the associated specificity 0.77 and sensitivity 0.80 (Fig. 1d). The accuracy of the four models was additionally checked using calibration plots (Fig. 2). The calibration plots for the third and fourth model show good calibration, since most of the points were close to the straight line (Fig. ...
Citations
... Grinter et al. (2019) validated MooMonitor's accuracy by comparing collected data with visual observation, confirming the device's accuracy in studies on cow behaviour [63]. In Borghart et al. (2021), the MooMonitor was used to predict lameness in cows through behavioural and production data, developing a predictive model combining accelerometer data with production parameters [64]. Krpálková et al. (2022) explored the correlation between rumination time detected by the collar and milk production at different lactation stages, noting that cows with longer rumination tended to produce more milk [65]. ...
Wearable collar technologies have become integral to the advancement of precision livestock farming, revolutionizing how dairy cattle are monitored in terms of their behaviour, health status, and productivity. These devices leverage cutting-edge sensors, including accelerometers, RFID tags, GPS receivers, microphones, gyroscopes, and magnetometers, to provide non-invasive, real-time insights that enhance animal welfare, optimize resource use, and support decision-making processes in livestock management. This systematized review focuses on analyzing the sensors integrated into collar-based systems, detailing their functionalities and applications. However, significant challenges remain, including the high energy consumption of some sensors, the need for frequent recharging, and limited parameter coverage by individual devices. Future developments must focus on integrating multiple sensor types into unified systems to provide comprehensive data on animal behaviour, health, and environmental interactions. Additionally, advancements in energy-efficient designs, longer battery life, and cost-reduction strategies are essential to enhance the practicality and accessibility of these technologies. By addressing these challenges, wearable collar systems can play a pivotal role in promoting sustainable, efficient, and responsible livestock farming, aligning with global goals for environmental and economic sustainability. This paper underscores the transformative potential of wearable collar technologies in reshaping the livestock industry and driving the adoption of innovative farming practices worldwide.
... In intensive dairy cattle farming, characterized by limited labor and increasingly large herd size, this approach becomes impractical (Clay et al., 2020). Furthermore, its time-intensive nature (Borghart et al., 2021) might result in an underestimation of lame cows (Alawneh et al., 2012;Dutton-Regester et al., 2020). Secondly, the traditional method is subjective as lameness scores are highly dependent on the observers' competence (Telezhenko and Bergsten, 2005;Renn et al., 2014;Kang et al., 2021). ...
This research paper proposes a simple image processing technique for automatic lameness detection in dairy cows under farm conditions. Seventy-five cows were selected from a dairy farm and visually assessed for a reference/real lameness score (RLS) as they left the milking parlor, while simultaneously being video-captured. The method employed a designated walking path and video recordings processed through image analysis to derive a new computerized automatic lameness score (ALDS) based on calculated factors from back arch posture. The proposed automatic lameness detection system was calibrated using 12 cows, and the remaining 63 were used to evaluate the diagnostic characteristics of the ALDS. The agreement and correlation between ALDS and RLS were investigated. ALDS demonstrated high diagnostic accuracy with 100% sensitivity and specificity and was found to be 100% accurate with a perfect agreement (ρ c = 1) and strong correlation ( r = 1, P < 0.001) for lameness detection in binary scores (lame/non-lame). Moreover, the ALDS had a strong agreement (ρ c = 0.885) and was highly correlated ( r = 0.840; 0.796–1.000 95% confidence interval, P < 0.001) with RLS in ordinal scores (lameness severity; LS1 to LS5). Our findings suggest that the proposed method has the potential to compete with vision-based lameness detection methods in dairy cows in farm conditions.
... A scalable decision tree boosting system (XGBoost) was selected as the classifier algorithm producing an Area under the Receiver Operating Characteristic Curve (AUC) of 86%, 81% SE, 78% SP and 81% F-measure (the harmonic mean of precision and recall). Borghart et al. (2021) evaluated predictive models gathering data from sensors and animal records to automatically detect lameness. Cows (n= 164) equipped with commercial neck-mounted 3D accelerometers (MooMonitor+, Dairymaster, Ireland) recording behavioural, rumination, resting and feeding time data, were used with data gathered over an 11-month period. ...
... The main reason for the popularity of these systems is that they involve a continuous sampling of the lameness predictor. Generally, lameness is detected using inertial measurements (Borghart et al., 2021;Jarchi et al., 2021;Weigele et al., 2018;Barker et al., 2018 and others), milk related measurements (Lemmens et al., 2023;Borghart et al., 2021;Van Hertem et al., 2016 and others), behavior-related predictors such as lying time, number of lying bouts, maximum length of the lying bout, roughage feeding time, etc. (Frondelius et al., 2022;Zhao et al., 2018;Thorup et al., 2016 and others) and a mixture of these mentioned predictors (Riaboff et al., 2021;Shahinfar et al., 2021 and others). ...
... The main reason for the popularity of these systems is that they involve a continuous sampling of the lameness predictor. Generally, lameness is detected using inertial measurements (Borghart et al., 2021;Jarchi et al., 2021;Weigele et al., 2018;Barker et al., 2018 and others), milk related measurements (Lemmens et al., 2023;Borghart et al., 2021;Van Hertem et al., 2016 and others), behavior-related predictors such as lying time, number of lying bouts, maximum length of the lying bout, roughage feeding time, etc. (Frondelius et al., 2022;Zhao et al., 2018;Thorup et al., 2016 and others) and a mixture of these mentioned predictors (Riaboff et al., 2021;Shahinfar et al., 2021 and others). ...
... This comparison would be valid for image/video -based studies as well as for the sensor-based studies, in which the number of observations/files are given. Jiang et al. (2022), Kang et al. (2020), Zhao et al. (2018), Lemmens et al. (2023), Borghart et al. (2021 are example of such studies which hold information needed for comparison. However, this comparison is not valid for the Frondelius et al. (2022), Antanaitis et al. (2021) and similar studies. ...
Lameness is one of the costliest pathological problems affecting dairy animals. It is usually assessed by trained veterinary clinicians who observe features such as gait symmetry or gait parameters as step counts in real time. With the development of artificial intelligence, various modular systems have been proposed to minimize subjectivity in lameness assessment. However, the major limitation in their development is the unavailability of a public database, as most existing ones are either commercial or privately held. To tackle this limitation, we have introduced CowScreeningDB, a multi-sensor database which was built with data from 43 dairy cows. Cows were monitored using smart watches during their normal daily routine. The uniqueness of the database lies in its data collection environment, sampling methodology, detailed sensor information, and the applications used for data conversion and storage, which ensure transparency and replicability. This data transparency makes Cow-ScreeningDB a valuable and objectively comparable resource for further development of techniques for lameness detection for dairy cows. In addition to publicly sharing the database, we present a machine learning technique which classifies cows as healthy or lame by using raw sensory data. To facilitate fair comparisons with state-of-the-art methods, we introduce a novel benchmark. Combining the database, the machine learning technique and the benchmark validate our major objective, which is to establish the relationship between sensor data and lameness. The developed technique reports an average accuracy of 77 % for the best case scenario and presents perspectives for further development. By introducing this framework which encompasses the database, the classification algorithm and the benchmark, we significantly reduce subjectively in lameness assessment. This contribution to lameness detection fosters innovation in the field and promotes transparent, reproducible research in the pursuit of more effective management of dairy cow lameness. Implications: Lameness detection is one of the main tasks in dairy systems, given its importance in the production ambit. However, the data used during detection is generally either held privately or sold commercially. In this study, we create a multi-sensor database (CowScreeningDB), which can be used for lameness. Because we have made the database public 1 and free of charge for research purposes, it should act as a benchmark allowing to objectively compare techniques put forth to deal with lameness. We also provide details of the sampling system used, comprised of hardware and a baseline classification algorithm.
... These new technologies provide a constant flow of high-frequency repeated measures for parameters such as milk yield and cow's activity which have shown to be sensitive of changes in the physiological and health status of the animal (Rutten et al. 2013;King and DeVries 2018). Recent models have been proposed for predicting lameness in dairy cattle based on automatically recorded data on cows' behavioural metrics and milk yield (O'Leary et al. 2020;Borghart et al. 2021). Moreover, recent innovative, cost-effective and rapid approaches for identifying and predicting lameness incidence at cow level are also increasingly being applied based on phenotyping technologies such as MIR spectrometry (Bonfatti et al. 2020;Contla Hern andez et al. 2021) or machine learning predictive algorithms (Warner et al. 2020;Shahinfar et al. 2021) that use routinely measured production and behavioural traits on farms. ...
... Several studies, however, have reported moderate to good inter-and intra-observer agreement for visual locomotion scoring [50][51][52][53]. Automated lameness detection such as the use of accelerometery, force pressure platforms and visionbased methods including video analysis and image processing have been evaluated [54][55][56]. The overall aim of automated lameness detection methods is to promptly identify and treat lame cows which have been reported to be associated with reduced duration and prevalence of lameness and improved production and welfare outcomes [10,11]. ...
Background
Lameness is a significant problem for the dairy industry worldwide. No previous studies have evaluated the prevalence of lameness or digital dermatitis (DD) in dairy cattle herds in Egypt. A total of 16,098 dairy cows from 55 dairy herds in 11 Egyptian governorates underwent visual locomotion scoring using a 4-point scoring system. Cows that had a lameness score ≥ 2 were considered clinically lame. Following manure removal with water and using a flashlight, the cows’ hind feet were examined in the milking parlour to identify DD lesions and classify with M-score. Furthermore, each cow was assigned a hock score (a 3-point scale) and a hygiene score (a 4-point scale). The cow-, within-and between-herd prevalence of lameness and DD and associated 95% confidence intervals (CI) were calculated. The prevalence of hock lesions and poor cow hygiene was also calculated.
Results
Of the examined cows, 6,883 were found to be clinically lame (42.8%, 95% CI = 42.0–43.5%). The average within-herd prevalence of lameness was 43.1% (95% CI = 35.9–50.3%). None of the dairy herds recruited into the study were found to be free from clinical lameness. The average within-herd prevalence of DD was 6.4% (95% CI = 4.9–8.0%). The herd-level prevalence of DD was 92.7% (95% CI = 85.9–99.6%). Active DD lesions (M1, M2, M4.1) were identified in 464 cows (2.9%) while inactive lesions (M3, M4) were identified in 559 cows (3.5%). The within-herd prevalence of hock lesions (score 2 or 3) was 12.6% (95% CI = 4.03–21.1%) while a severe hock lesion had within-herd prevalence of 0.31% (95% CI = 0.12–0.51%). Cow-level prevalence of hock lesions was 6.2% (n = 847, 95% CI = 5.8–6.2%). The majority of examined cows had a hygiene score of 4 (n = 10,814, prevalence = 70.3%, 95% CI = 69.5–71%).
Conclusions
The prevalence of lameness was higher than prevalence estimates reported for other countries which could be due to differing management and/or environmental factors. DD was identified at lower prevalence in most herds but with high herd-level prevalence. Poor cow hygiene was notable in most herds. Measures to reduce the prevalence of lameness and to improve cow hygiene in dairy cattle herds in Egypt are therefore needed.
... Digitalization and automatization have advanced rapidly in the agricultural and livestock sector over the past decades [1][2][3] benefiting both farmers and animals. Further, the use of automated milking systems (AMSs) has increased globally within the dairy industry, automatically and continuously providing farmers with various data outputs concerning milk yield, milk quality, and animal activity [4][5][6][7]. ...
... In contrast, other research approaches focus on automated lameness detection using devices already available on dairy farms. For example, sensors which are directly attached to the cow are frequently used for health monitoring nowadays, by recording behavioral patterns such as rumination, eating, or lying down, relying on the change of these movements for disease detection [2,16,43,44]. Various studies revealed a significant effect of lameness on the cows' eating time, i.e., the time to secure sufficient dry matter intake, daily milk yield, the number of daily visits to the AMS, and the number of cows that had to be fetched to the AMS, resulting in a higher workload [4,5,45]. A study from Miguel-Pacheco et al. [46] carried out in farms equipped with an AMS showed a significant negative association between total feeding time and lameness as well as the frequency of feeding bouts and lameness. ...
... Lameness detection solely relying on sensor parameters resulted in the lowest accuracy of 0.623 (±0.005) with a sensitivity of 0.610 (±0.009) and a specificity of 0.640 (±0.030). A study by Borghart et al. [2] found similar results for lameness detection using parameters from a different sensor system (MooMonitor+ ® , Dairymaster ® , Causeway, Ireland) with our model outperforming in terms of specificity (64% versus 53%). ...
Simple Summary
The objective of the present study was to develop a tool to detect mildly lame cows by combining already existing data from sensors, automated milking systems (AMSs), routinely recorded animal and farm data and other phenotypes. Ten dairy farms were visited every 30–42 days from January 2020 to May 2021, and locomotion scores (LCSs) and body condition scores (BCSs) were assessed at each visit. For each farm, a lameness incidence risk was calculated. Further, the impact of lameness on the derived sensor parameters was inspected. Finally, random forest models for lameness detection were fit by including different combinations of influencing variables and compared for best results. The best performing model achieved an accuracy of 0.75 with a sensitivity of 0.72 and specificity of 0.78 for predicting an LCS ≥ 2. We conclude that the combination of data from an available sensor device with routinely available AMS data, animal and farm information, and performance records can provide promising results for the detection of mild lameness in dairy cattle.
Abstract
This study aimed to develop a tool to detect mildly lame cows by combining already existing data from sensors, AMSs, and routinely recorded animal and farm data. For this purpose, ten dairy farms were visited every 30–42 days from January 2020 to May 2021. Locomotion scores (LCS, from one for nonlame to five for severely lame) and body condition scores (BCS) were assessed at each visit, resulting in a total of 594 recorded animals. A questionnaire about farm management and husbandry was completed for the inclusion of potential risk factors. A lameness incidence risk (LCS ≥ 2) was calculated and varied widely between farms with a range from 27.07 to 65.52%. Moreover, the impact of lameness on the derived sensor parameters was inspected and showed no significant impact of lameness on total rumination time. Behavioral patterns for eating, low activity, and medium activity differed significantly in lame cows compared to nonlame cows. Finally, random forest models for lameness detection were fit by including different combinations of influencing variables. The results of these models were compared according to accuracy, sensitivity, and specificity. The best performing model achieved an accuracy of 0.75 with a sensitivity of 0.72 and specificity of 0.78. These approaches with routinely available data and sensor data can deliver promising results for early lameness detection in dairy cattle. While experimental automated lameness detection systems have achieved improved predictive results, the benefit of this presented approach is that it uses results from existing, routinely recorded, and therefore widely available data.
... These new technologies provide a constant flow of high-frequency repeated measures for parameters such as milk yield and cow's activity which have shown to be sensitive of changes in the physiological and health status of the animal (Rutten et al. 2013;King and DeVries 2018). Recent models have been proposed for predicting lameness in dairy cattle based on automatically recorded data on cows' behavioural metrics and milk yield (O'Leary et al. 2020;Borghart et al. 2021). Moreover, recent innovative, cost-effective and rapid approaches for identifying and predicting lameness incidence at cow level are also increasingly being applied based on phenotyping technologies such as MIR spectrometry (Bonfatti et al. 2020;Contla Hern andez et al. 2021) or machine learning predictive algorithms (Warner et al. 2020;Shahinfar et al. 2021) that use routinely measured production and behavioural traits on farms. ...
... Sensor technology can provide valuable information on animal health, behaviour and welfare without the need for visual observations (Alvarenga et al. 2016; Barwick et al. 2018;Walton et al. 2018; Barwick et al. 2020). Monitoring behaviour of ruminants can be a reliable tool for predicting lameness (Borghart et al. 2021), detecting oestrus (Kamphuis et al. 2012;O'Neill et al. 2014), identifying the mother (Sohi et al. 2017), evaluating gastrointestinal nematode infection (Ikurior et al. 2020), predicting lambing time (Fogarty et al. 2020a(Fogarty et al. , 2020bSmith et al. 2020;Gurule et al. 2021) and predicting calving time (Borchers et al. 2017). Therefore, the wearable sensors can be used to facilitate management of farms as well as increase their profitability. ...
Context
Lamb loss and dyctocia are two major challenges in extensive farming systems. While visual observation can be impractical due to the large sizes of paddocks, number of animals and high labour cost, wearable sensors can be used to monitor the behaviour of ewes as there might be changes in their activities prior to lambing. This provides sufficient time for the farm manager to nurse those ewes that are at risk of dyctocia.
Aim
The objective of this study was to determine whether the behaviour of a pregnant ewe could predict the time of parturition.
Methods
Two separate trials were conducted: the first trial (T1), with 32 ewes, included human/video observations, and the second trial (T2), with 165 ewes, conducted with no humans present, to emulate real extensive farming settings. The ewes were fitted with tri-axial accelerometer sensors by means of halters. Three-dimensional movement data were collected for a period of at least 7 and 14 days in T1 and T2 respectively. The sensor units were retrieved, and their data downloaded using ActiGraph software. Ewe behaviour was determined through support vector machine learning (SVM) algorithm, including licking, grazing, rumination, walking, and idling. The behaviours of ewes predicted by analysis of sensor data were compared with behaviours determined using visual observation (video recordings), with time synchronisation to validate the results. Deep learning and neural-network algorithms were used to predict lambing time.
Key results
The concordance percentages between visual observation and sensor data were 90 ± 11, 81 ± 15, 95 ± 10, 96 ± 6, and 93 ± 8% ± s.d. for grazing, licking, rumination, idling, and walking respectively. The deep-learning model predicted the time of lambing with 90% confidence via a quantile regression method, which can be interpereted as 90% prediction intervals, and shows that the time of lambing can be predicted with reasonable confidence approximately 240 h before the actual lambing events.
Conclusion
It was possible to predict the time of parturition up to 10 days before lambing.
Implications
The behaviour of ewes around lambing time has a direct effect on the survival of the lambs and therefore plays an important part in animal management. This knowledge could improve the productivity of sheep and considerably decrease lamb mortality rates.
... Sensor technology can provide valuable information on animal health, behaviour and welfare without the need for visual observations (Alvarenga et al. 2016; Barwick et al. 2018;Walton et al. 2018; Barwick et al. 2020). Monitoring behaviour of ruminants can be a reliable tool for predicting lameness (Borghart et al. 2021), detecting oestrus (Kamphuis et al. 2012;O'Neill et al. 2014), identifying the mother (Sohi et al. 2017), evaluating gastrointestinal nematode infection (Ikurior et al. 2020), predicting lambing time (Fogarty et al. 2020a(Fogarty et al. , 2020bSmith et al. 2020;Gurule et al. 2021) and predicting calving time (Borchers et al. 2017). Therefore, the wearable sensors can be used to facilitate management of farms as well as increase their profitability. ...
Context. Lamb loss and dyctocia are two major challenges in extensive farming systems. While visual observation can be impractical due to the large sizes of paddocks, number of animals and high labour cost, wearable sensors can be used to monitor the behaviour of ewes as there might be changes in their activities prior to lambing. This provides sufficient time for the farm manager to nurse those ewes that are at risk of dyctocia. Aim. The objective of this study was to determine whether the behaviour of a pregnant ewe could predict the time of parturition. Methods. Two separate trials were conducted: the first trial (T1), with 32 ewes, included human/video observations, and the second trial (T2), with 165 ewes, conducted with no humans present, to emulate real extensive farming settings. The ewes were fitted with tri-axial accelerometer sensors by means of halters. Three-dimensional movement data were collected for a period of at least 7 and 14 days in T1 and T2, respectively. The sensor units were retrieved, and their data downloaded using ActiGraph software. Ewe behaviour was determined through Support Vector Machine Learning (SVM) algorithm including licking, grazing, rumination, walking, and idling. The behaviours of ewes predicted by analysis of sensor data were compared with behaviours determined using visual observation (video recordings), with time synchronisation to validate the results. Deep learning and neural network algorithms were used to predict lambing time. Key results. The concordance percentage between visual observation and sensor data were 90±11, 81±15, 95±10, 96±6, and 93±8 percent±SD for grazing, licking, rumination, idling, and walking, respectively. The deep learning model predicted the time of lambing with 90% confidence via a quantile regression method, which can be interpereted as 90% prediction intervals, and shows that the time of lambing can be predicted with reasonable confidence approximately 240 hours before actual lambing events. Conclusion. It was possible to predict the time of parturition up to 10 days before lambing. Implications. The behaviour of ewes around lambing time has a direct effect on the survival of the lambs and therefore plays an important part in animal management. This knowledge could improve the productivity of sheep and considerably decrease lamb mortality rates.
